Developers building AI features
Connect a compatible SDK, test models in the playground, and review the requests behind your application. Change virtual-model targets without changing the model selector in your code.
ABOUT ZROUTER
Not every request
needs your biggest model.
ZRouter is a Super Intelligence Router. It reads each request and sends it to the model that fits: quick answers to fast, low-cost models and hard reasoning to your strongest one. One OpenAI-compatible endpoint, every provider behind it, and keys, usage, and credits in one workspace.
Keep model choices open.
Keep the work moving.
Adding an AI feature should not mean rebuilding your integration every time you try a different provider. Managing that feature should not require hunting through separate keys, usage reports, and billing records.
ZRouter brings those tasks together. Connect an OpenAI-compatible client, choose an available model, and use your workspace to manage access and follow usage. A private virtual model gives your application a stable name while you adjust the targets behind it.
Our purpose is practical: help you spend more time building useful AI features and less time managing their connections.
WHO IT IS FOR
From a first prototype
to an everyday workflow.
For people who want model choice and account controls close to the work they already do.
People managing AI usage
Give each application its own key, review token activity and settled charges, and manage account budgets and rate limits when enabled. Use prepaid credits or supported provider keys.
Developers working with agents
Use zctl in your terminal or connect an MCP client such as Claude Code or Cursor. Inspect account usage, manage routers, and work with keys and limits where you already build.
WORK WHERE YOU BUILD
One account.
Three ways to manage it.
Use the dashboard for a visual view, zctl for repeatable commands, or MCP to work with your AI assistant.
See it in the dashboard
Compare usage, inspect available request logs, manage credits, and test models in the playground. Start here when you want to see the whole account.
Open the quickstart ↗Automate it with zctl
Create and edit virtual routers, manage keys, and review usage from your terminal. JSON output and router exports make account tasks easier to repeat in scripts.
Install the zctl CLI ↗Ask your agent through MCP
Connect Claude Code, Cursor, or another compatible MCP client. Ask about usage, investigate errors, and manage routers, API keys, and enabled account limits.
Connect an MCP client ↗CLI and MCP management access is limited to your account. Connecting MCP does not change your agent's inference provider. See each tool's integration guide for model-request support.
THE PRINCIPLES BEHIND THE PLATFORM
Simple where it counts.
Explicit where it matters.
These principles guide the product and how we explain its capabilities.
A familiar connection
Use OpenAI-compatible requests across supported providers. Check the API guide for endpoint and model capabilities before connecting an application.
Choices you can change
Choose provider models or create private virtual models. Keep the selector in your application stable as your routing needs evolve.
Credentials you control
Saved BYOK secrets are encrypted at rest. Create and revoke your account API keys and management tokens. CLI and MCP access stays scoped to your own account.
Costs you can inspect
New accounts start at $0 with no monthly subscription. Credit purchases start at $5. Your ledger records settled charges, and BYOK provider fees remain separate from ZRouter routing fees.
Developed and maintained
by NEXUS AI.
NEXUS AI develops and maintains ZRouter. Our focus here is a practical workspace for connecting applications to AI models and managing their usage.
Learn about the organization behind the product or contact the team with a product question.
ABOUT THE PRODUCT
Good questions.
Clear starting points.
What is ZRouter?
ZRouter is an AI gateway and account workspace. It routes supported requests to model providers through an OpenAI-compatible API and brings model selection, keys, usage, logs, and prepaid credits together.
Why use ZRouter instead of managing each connection separately?
Use one compatible integration and one place to manage keys, routing, and usage. Virtual models let you change provider targets behind a stable selector, and the dashboard, zctl CLI, and MCP server give you different ways to manage the same account.
Does MCP route my coding agent's model requests?
The account MCP server lets an agent manage your ZRouter account. It does not change the agent's model provider or enable tool-based inference on prepaid accounts. The MCP guide explains management access; the integration guides explain model connections and their requirements.
YOUR NEXT REQUEST STARTS HERE
Build with AI.
Keep the controls.
One account for model access, routing, and usage. Manage it from your dashboard, terminal, or AI assistant.